eduzhai > Applied Sciences > Transportation >

Statewide Modeling of Pedestrian and Bicyclist Safety

  • Save

... pages left unread,continue reading

Document pages: 16 pages

Abstract: Many cities and metropolitan planning organizations devote considerable resources to addressing pedestrian crash risk given its importance, particularly in urban areas. Most of these entities, however, lack the requisite toolkit to conduct such an analysis, because relying merely on hotspots or observed crashes for generating these insights does not take into consideration pedestrian exposure. We address exposure by building trip generation tables at fine geographic scale via synthetic households using iterative proportional fitting method. This approach allows us to carry out the study statewide while preserving the strength of disaggregated studies. The inputs to the model include US Census level microdata, a statewide travel survey and built environment data. The output is a measure of pedestrian risk at the pedestrian analysis zone (PAZ) level of 400 meters by 400 meters.In generating the pedestrian risk estimates, we utilize an empirical Bayes approach to define a pedestrian risk score based not only on the spatial patterns of crashes but also on a multitude of factors including daily traffic and pedestrian exposure to crashes. In so doing, we develop a framework to estimate pedestrian exposure to risk at a very fine geographical scale – the pedestrian analysis zone (PAZ). This provision addresses a shortcoming of the safety performance functions (SPFs) in the Highway Safety Manual for pedestrian-vehicle crashes, as it does not include a measure of pedestrian volume. Agencies can use our approach to generate risk scores to estimate pedestrian safety and identify zones or facilities where countermeasures need to be applied.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...